Caving Headlamp – Beam Shapes
By Ian Mander 24-27 January 2010, updated 26-30 October, 5 & 8 November 2010, 24 June 2011.
Using multiple LEDs allows the use of different optics, providing greater versatility of beam shapes if the LEDs are suitably switched. To get the LEDs into the smallest possible space I'll be using 10 mm optics, which are reasonably inexpensive, reasonably efficient, and available in several beam shapes.
Carclo footnote: Please note that flow lines, weld lines, surface scratches and small black or white inclusions within the lenses are acceptable if the optical performance of the lens is within the specification described to ±10%.
As typical for Cutter's pathetic site, they appear to get important details wrong, wrong, wrong, which is particularly sad if people are basing their buying decisions on the Cutter-provided data.
Wide Beam – General Use
For general caving I prefer a centre-weighted flood. Unfortunately I haven't really figured out what the best beam width or centre weighting is, or even if there is an optimum combination. It probably depends on what type of caving I'm doing, and I suspect cavers generally tend to make do with whatever they're using because it's actually quite tricky to compare or test such things.
What I am convinced of is that no centre weighting at all – a pure flood – is very undesirable because it means no distance viewing ability at all, which is actually quite annoying. Also, light is basically wasted on my low-detail peripheral vision instead of being ahead, where I'm looking, and where my detail vision needs the light. This rules out naked LEDs – the amount of light required to get reasonable distance is too high. Even the XR-E, with its 90° FWHM beam angle (narrower than typical) is still too wide.
However, all centre spot is also unpleasant. It results in tunnel vision and more head movement than desirable. In tiny passages the intensity of the central spot on a nearby wall can be dazzling.
The beam pattern of 5 mm LED arrays I've used is quite pleasant. The centre spot isn't sharply defined, and it's diffuse enough that (a) my pupils don't stop down for just the brightest part and (b) there's enough side spill to keep my peripheral vision happy.
The two leading options for the wide beam optics are the frosted medium at 30° and the frosted wide at 40°. I'm strongly tending more toward the frosted medium.
Spot Beam – Distant Viewing
Spotlighting objects on the far side of chambers, down holes, or up avens is very useful. Thus, any reasonable caving headlamp should have at least moderate distant lighting performance, also known as throw.
I'll note here that warm white LEDs may throw better in some conditions such as humid caves, because yellow light scatters less than blue light so will result in less glare, especially back-scatter. My personal feeling from experience with car headlights is that even though we're told to drive on low beam in fog, and high beam certainly provides more glare than low beam, it actually gives better visibility as well. The lower efficacy of warm white LEDs almost certainly doesn't make up for the lower amount of glare (although they may simply be much more pleasant to use). However, neutral white LEDs are available up to R4 bin for the XP-G and T5 or T6 for the XM-L, and will provide many of the colour advantages of warm white without the yellowish colour and relative inefficiency.
For extreme throw an aspheric lens works best. Even though a fair bit of light is lost sideways without being captured by the lens, the light that is captured is so well focused that it illuminates objects further away much better than an optic or reflector does. (With the possible exception of a catadioptic reflector. That's another story.)
The aspheric lens, of course, influences the beam. The wider the lens the more light it will capture. The longer the focal length of the lens the tighter the focused beam will be, but the less light it will capture because it's further from the LED. The bigger the lens the heavier it is, and the less desirable for caving.
The LED also influences the beam. For the same brightness, the smaller the LED die the tighter the beam because the light is being produced from a smaller area, which means an XP-E will produce a tighter beam than an XP-G even though they have the same LED package. Because the beam is tighter, an XP-E R2 will throw further than a more efficient XP-G R5 at the same current, even though the XP-E is producing less total light. However, the XP-E is only rated up to 1 A while the XP-G is rated up to 1.5 A, so on full current the XP-G will probably throw a brighter and wider beam.
The XR-E R2 will produce roughly the same amount of light as the XP-E R2 at low currents, but has the edge at higher current because the XR-E package is better able to cope with heat. Another advantage is its beam angle (without any secondary optics) is narrower than the XP-E, which means more of its output light would be collected by an aspheric lens centrally placed in front of it. The downside of the XR-E is that it produces a ring of light around the die projection – a cosmetic problem only.
Basically, a case could be made for any of the LEDs in the table, possibly excepting the XM-L (which will likely throw a very wide beam with any practically sized aspheric lens). They would all do well and the final choice might come down to what's easily and inexpensively available, rather than on what sort of beam is wanted.
Having said all that, the simplest way of getting a caving spot beam is to use a 3xAAA C30 zoom head torch mounted on the side of the helmet. It uses an XR-E direct driven at anything up to 1.3 A, and gives a brighter spot than the 1xAA C78 zoom head torch. Its versatility (in being able to zoom) is also a strong advantage and is also attractive in providing a backup helmet-mounted light.
Because the side-mounted zoom head torch will be filling the role of aspheric thrower I'm not going to have that function in the new headlamp. I prefer to have the LEDs fitted with more general use optics - which an aspherical lens is not - so they will be used more of the time.
For a tighter beam in the new headlamp, the Carclo 10 mm 10412 plain tight optic's beam is a little too die-shaped, so the 10417 lens front narrow optic looks the best option, providing a 24° beam. This is quite wide compared to many optics (eg, 8° from the standard XR-E TIR optic), but it's hard to get a very narrow beam from an XP-G.
Whether this proves to be a good idea will be tested in the field (cave). Update: With an XP-G at high current it throws a bright "wall" of light. Impressive.